基于 Stacked ConvLSTM 的时间序列森林火烧迹地检测

As the largest land cover, forest plays an important role in human living environment, biological habitat and global carbon cycle. Forest health is directly related to global ecological security and sustainable development of human society. In recent years, urban construction, disasters, forest management and deforestation and other factors have caused different degrees of disturbance to forests. It is important to determine the exact time point and spatial range of forest burned area for forest damage assessment, management, carbon accounting and forest restoration management. Because of the continuity of spatial distribution of forest burned areas, most of the existing methods of forest burned areas extraction use the two-step treatment strategy of first classification and then post-processing to suppress the effect of false alarm pixels. In this paper, a spatiotemporal detection method, Stacked ConvLSTM, is proposed for the detection of forest fire tracks in time series, which avoids subjective post-processing operations on the basis of maintaining better spatial continuity of the results, and achieves end-to-end extraction of forest burned areas information, which improves the extraction accuracy of forest fire-burning land. (Objective)This paper proposes to use Stacked ConvLSTM to detect forest disturbance in time and space. Combined with the characteristics of ConvLSTM in extracting temporal and spatial characteristics from long-term historical series, it can predict the change trend of vegetation in a period of time in the future, and more accurately determine the time point and spatial range of forest disturbance. (Method) ConvLSTM is a LSTM variant proposed on the basis of LSTM. The full connection state from input layer to hidden layer and from hidden layer to hidden layer of LSTM is replaced by convolution connection, which can make full use of spatial information. Compared single pixel based methods, ConvLSTM can extract the spatiotemporal structure information of time series images at the same time, better for spatiotemporal analysis. In this paper, Stacked ConvLSTM is used to detect the temporal and spatial distribution of forest burned areas, predict the change trend of vegetation in a period of time in the future, and determine whether there is forest burned areas by comparing with the newest time-series images.(Result) With MODIS long time series data, based on the historical time series of Yinanhe Forest Farm of Zhanhe Forestry Bureau in Heilongjiang Province and Beidahe Forest Farm of Bilahe Forestry Bureau in Inner Mongolia from 2001 to 2008 and 2001-2016, the extraction results of burned areas were compared with Stacked LSTM and bfast algorithm. The Stacked ConvLSTM, Stacked LSTM, and bfast algorithms were used to extract forest burned areas from MODIS time series in both regions, and compare the detection results with the Fire_CCI 5.1 burned areas products released by ESA. The results show that, first of all, from the visual effect, In Study Area I, the error detection of Stacked ConvLSTM in the study area are fewer than Stacked LSTM and bfast algorithm and maintain high continuity in spatial distribution. In Study Area II, Stacked ConvLSTM detected a more complete area of fire. Secondly, in the first study area, Stacked ConvLSTM was 0.120 and 0.405 more accurate than Stacked LSTM and bfast algorithms respectively, and the recall rate, accuracy, and Fire_CCI F1-score were higher. In the second study area, the accuracy of Stacked ConvLSTM is 0.924, with a higher recall rate, accuracy, and F1-score than Stacked LSTM and bfast algorithms and Fire_CCI 5.1. (Conclusion) The detection accuracy of ConvLSTM model in space is higher than the other two methods, and the continuity of detection results in space is better. The detection effect of ConvLSTM model is equivalent to that of Stacked LSTM in time, but both of them are closer to the real fire time point than bfast algorithm. The results show that Stacked ConvLSTM has certain advantages in obtaining the change trend of forest long-term historical series for spatiotemporal prediction, and improves the detection accuracy of forest fire to a certain extent.